Service provider management of virtual instances corresponding to hardware resources managed by other service providers

• US 9,213,503 B2
• Filed: 10/30/2011
• Issued: 12/15/2015
• Est. Priority Date: 10/30/2011
• Status: Expired due to Fees

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: B, Kamath Harish, Rajesh, Bhatia
• Primary Examiner(s): Cao, Diem
• Application Number:
  US13/284,944
• Publication Number:
  US 20130111469A1
• Time in Patent Office: 1,507 Days
• Field of Search: 718/105
• US Class Current: 1/1
• CPC Class Codes: G06F 2206/1508 Load balancing
• CPC Class Codes: G06F 3/1211 Improving printing performance
• CPC Class Codes: G06F 3/124 Parallel printing or parallel ripping
• CPC Class Codes: G06F 3/1287 via internet
• CPC Class Codes: G06F 9/5044 considering hardware capabilities
• CPC Class Codes: G06F 9/505 considering the load

Abstract

A dynamic pool of virtual instances is managed. Each virtual instance corresponds to one or more hardware resources. Management includes tracking for each virtual instance: historical and current utilization, an operating cost, a capacity, and a characteristic associated with a capability of the hardware resources to which the virtual instance corresponds. A given virtual instance of the dynamic pool of virtual instances on which to schedule a workload for completion is selected, based on a characteristic of the workload matching the characteristic of the given virtual instance, and taking into account the capacity, the current utilization, and the operating cost of the given virtual instance. The hardware resources are managed by service providers different than a service provider managing computing devices on which the virtual instances are managed and the workload is scheduled, and the workload is received from a customer different than any of these service providers.

Claims

1. A system comprising:

one or more computing devices communicatively coupled to a plurality of hardware resources, the computing devices maanaged by a first service provider, the hardware resources managed by one or more second service providers different than the first service provider;

a pool manager to manage a dynamic pool of virtual instances, each virtual instance corresponding to one or more of the hardware resources, the pool manager to track for each virtual instance;

historical and current utilization, an operating cost, a capacity, and a characteristic associated with a capability of the virtual instance; and

a scheduler to select a given virtual instance of the dynamic pool of virtual instances on which to schedule a workload for completion, based on a characteristic of the workload matching the characteristic of the given virtual instance, and taking into account the capacity, the current utilization, and the operating cost of the given virtual instance,

wherein the pool manager and the scheduler are implemented on the computing devices, and wherein the scheduler is to;

receive a job from a customer, the customer being different than the first service provider and the one or more second service providers,

divide the job into a plurality of job components,

hierarchically organize the plurality of job components,

determine a workload for each of the plurality of job components,

assign the workloads to the virtual instances in the dynamic pool of virtual instances based on matching characteristics of the workloads with the characteristics of the virtual instances, and

schedule the workloads for completion on the assigned virtual instances.

2. The system of claim 1, wherein the pool manager is to add a new virtual instance to the dynamic pool based on tracking of the historical and current utilization, the capacity, and the characteristic of each virtual instance already in the dynamic pool.

3. The system of claim 1, wherein the pool manager is to cause or request the capacity of an existing virtual instance of the dynamic pool to be augmented with additional of the hardware resources based on tracking of the historical and current utilization, the capacity, and the characteristic of at least the existing virtual instance in the dynamic pool.

4. The system of claim 1, wherein the pool manager is to tag an existing virtual instance of the dynamic pool for removal from the dynamic pool based on tracking of the historical and current utilization, the capacity, and the characteristic of each virtual instance already in the dynamic pool.

5. The system of claim 1, wherein the pool manager is to remove an existing virtual instance from the dynamic pool upon the existing virtual instance completing uncompleted workloads already scheduled on the existing virtual instance.

6. The system of claim 1, wherein, when a given workload is not completed by one of the virtual instances, the scheduler is to duplicate the given workload and assign the duplicate workload to another virtual instance in the dynamic pool of virtual instances.

7. The system of claim 1, wherein a first hierarchically organized job component is a parent job component to a second hierarchically organized job component, and the workload corresponding to the second hierarchically organized job component loosely inherits one or more characteristics from the workload corresponding to the first hierarchically organized job component.

8. The system of claim 1, wherein a first hierarchically organized job component is a parent job component to a second hierarchically organized job component, and the workload corresponding to the second hierarchically organized job component strictly inherits one or more characteristics from the workload corresponding to the first hierarchically organized job component.

9. The system of claim 1, wherein a first hierarchically organized job component is a parent job component to a second hierarchically organized job component, and the workload corresponding to the first hierarchically organized job component is scheduled on a first virtual instance of the dynamic pool that is different than a second virtual instance of the dynamic pool on which the workload corresponding to the second hierarchically organized job component is scheduled.

10. A non-transitory computer-readable data storage medium storing a computer program executable by a computing device to cause the computing device to:

manage a dynamic pool of virtual instances, each virtual instance corresponding to one or more hardware resources of a plurality of hardware resources to which the computing device is communicatively coupled, wherein the computing device is to track characteristics of each virtual instance including;

historical and current utilization, an operating cost, a capacity, and a characteristic associated with a capability of the virtual instance,

and wherein the computing device is managed by a first service provider, and the hardware resources are managed by one or more second service providers different than the first service provider;

receive a job from a customer, the customer being different than the first service provider and the one or more second service providers;

divide the job into a plurality of job components;

hierarchically organize the plurality of job components;

determine a workload for each of the plurality of job components;

assign the workloads to the virtual instances in the dynamic pool based on matching characteristics of the workloads with the characteristics of the virtual instances; and

schedule the workloads for completion on the assigned virtual instances.

11. The non-transitory computer-readable data storage medium of claim 10, wherein to manage the dynamic pool of virtual instances, the computer program is to cause the computing device to add a new virtual instance to the dynamic pool, based on tracking of the historical and current utilization, the capacity, and the characteristic of each virtual instance already in the dynamic pool.

12. The non-transitory computer-readable data storage medium of claim 10, wherein to manage the dynamic pool of virtual instances, the computer program is to cause the computing device to request the capacity of an existing virtual instance of the dynamic pool to be augmented with additional of the hardware resources based on tracking of the historical and current utilization, the capacity, and the characteristic of at least the existing virtual instance in the dynamic pool.

13. The non-transitory computer-readable data storage medium of claim 10, wherein to manage the dynamic pool of virtual instances, the computer program is to cause the computing device to remove an existing virtual instance from the dynamic pool, including tagging the existing virtual instance of the dynamic pool for removal from the dynamic pool based on tracking of the historical and current utilization, the capacity, and the characteristic of each virtual instance already in the dynamic pool.

14. The non-transitory computer-readable data storage medium of claim 10, wherein to manage the dynamic pool of virtual instances, the computer program is to cause the computing device to remove an existing virtual instance from the dynamic pool occurs upon the existing virtual instance completing uncompleted workloads already scheduled on the existing virtual instance.

15. A method comprising:

managing, by a computing device managed by a first service provider, a dynamic pool of virtual instances on which to schedule a workload for completion, each virtual instance corresponding to one or more hardware resources of a plurality of hardware resources to which the computing device is communicatively coupled, the hardware resources managed by one or more second service providers different than the first service provider,

receiving a job from a customer, the customer being different than the first service provider and the one or more second service providers;

dividing the job into a plurality of job components;

hierarchically organizing the plurality of job components;

determining a workload for each of the plurality of job components;

assigning the workloads to the virtual instances in the dynamic pool based on matching characteristics of the workloads with the characteristics of the virtual instances, taking into account the capacity, the current utilization, and the operating cost of the virtual instances; and

scheduling the workloads for completion on the assigned virtual instances.

16. The method of claim 15, further comprising:

when a given workload is not completed by one of the virtual instances, duplicating the given workload and assigning the duplicate workload to another virtual instance in the dynamic pool of virtual instances.

17. The method of claim 16, wherein a first hierarchically organized job component is a parent job component to a second hierarchically organized job component, and the workload corresponding to the second hierarchically organized job component strictly inherits one or more characteristics from the workload corresponding to the first hierarchically organized job component.

18. The method of claim 16, wherein a first hierarchically organized job component is a parent job component to a second hierarchically organized job component, and the workload corresponding to the first hierarchically organized job component is scheduled on a first virtual instance of the dynamic pool that is different than a second virtual instance of the dynamic pool on which the workload corresponding to the second hierarchically organized job component is scheduled.

19. The method of claim 15, wherein a first hierarchically organized job component is a parent job component to a second hierarchically organized job component, and the workload corresponding to the second hierarchically organized job component loosely inherits one or more characteristics from the workload corresponding to the first hierarchically organized job component.

20. The method of claim 15, further comprising:

removing, by the computing device managed by the first service provider, an existing virtual instance from the dynamic pool occurs upon the existing virtual instance completing uncompleted workloads already scheduled on the existing virtual instance.